An Efficient Approach to Ill-Posed Problem Regularization Applied to an Over-Determined Six-Degree of Freedom Load Cell

There is an increasing demand for the development of six-degree of freedom load cells, especially for the application in the field of sport equipments or for the identification of mechanical properties. The developed load cell makes it possible to measure the six loads (three forces and three moments) transmitted by a skier to one ski by eight Wheatstone bridge channels. The paper deals with the experimental procedure for calibration and with validation tests to check measurement accuracy, whose initial outcome was that the cell response was not sufficiently accurate and precise. Therefore, the regularization algorithm by Tikhonov-Phillips was applied to improve results. From the methodological point of view, a new criterion is proposed here for the selection of the regularization parameter (damping factor). This approach is based on the random generation of noisy signals and on the adoption of an error estimator that accounts for all the loads being simultaneously applied and measured. A significant improvement of the load cell accuracy could be achieved, even in presence of noise, moreover the proposed method is particularly efficient and not computationally expensive.